Dear memberships,
I'm trying to estimate the following multivariate local regression model using
the "locfit" package:
BMI=m1(RCC)+m2(WCC)
where (m1) and (m2) are unknown smooth functions.
My problem is that once I get the regression done I cannot get the fitted
values of each of this smoo
Dear memberships,
I'm trying to estimate the following multivariate local regression model using
the "locfit" package:
BMI=m1(RCC)+m2(WCC)
where (m1) and (m2) are unknown smooth functions.
My problem is that once I get the regression done I cannot get the fitted
values of each of this
Dear members,
I'm trying to do a local polynomial estimation where the weight function is the
inverse of the kernel (using a gaussian kernel). Does anybody know how I can do
it?
When I work with a weigth function usual I use:
locfif(y~lp(x,deg=1,h=0.75),kern="gauss")
Does anybody know
Dear members,
I'm trying to estimate a varying coefficients model using the local polynomial
estimation method in two case (univariate and bivariate) and with two smooth
functions.
In the univariate case I use:
m1<-smooth.lf(x=lp(z1,by=x1,deg=1,h=0.7,ev=z1)+lp(z3,by=expl2,deg=1,h=0.7,ev=z1
Hello dear members,
I need to calculate "by hand" a local lineal regression so I need to compute a
kernel weight.
Does somebody knows how to get a kernel to use as weighted? I can calculate a
density kernel function and after pre-multiply it by the sample size. However I
know this is not w
Hello to everybody,
Somebody know how I can apply the instruction (regband) of the package (locfit)
to a multivariate covariate. In the instructions states that the formula has to
have only one predictor.
How can I use that when I have an unknown function with two predictors, m(X,Z)?
Is there
Hello,
I trying to find a way to calculate a bivariate bandwidth for a bivariate local
polynomial regression and I have a question.
Somebody know if I can use the instruction "dpill" to do it?
When I read the instructions I see that "x" must be a vector of variables, but
if I state "x" is
Hi,
I want to calculate the bandwith of a multivariate local polynomial regression
such as the model is: Y=m(X,Z)+u.
Please, somebody knows how I can do it in a multivariate case? In an univariate
case I use the instrucction "regband" of the locfit package, but it is
impossible to find a
Dear all,
I'm trying to do a multivariate local regression whit the locfit instruction,
such as:Y=m(Z,X,W)+u
However, I have a problem at the moment of calculte the bandwith of the
regression. If I had a univarite local regression model I could use the
instruction "regband". However, t
Hello,
I want to create 2 variables with normal distribution that the correlation
between them is equal to 0.8 and also each one has a serial correlation equal
to 0.6.
To generate variables with correlation between them I use:
t1<-10
N<-45
sigma2<-matrix(c(1,0.8*sqrt(1),0.8*sqrt(1),1),
Hello,
I want to create two random variables (x1,x2) both with uniform distribution
bounded by (-1) and (1) that has a correlation of 0.6 between them.
Does somebody know how I can do it? For normal random variables I known how to
implement it with the rmvnorm command but I don't know how t
Hello,
I want to create a matrix of N random numbers with a uniform distributions.
Later, I want to repeat T times each row of this matrix. For this I do the
following loop:
N<-45
T<-10
n<-N*T
a<-matrix(runif(N,min=-1,max=1),nr=N)
mymat<-matrix(rep(NA,n),nr=n,nc=1)
for(i in i:N){
b<-rep(a[i,],
12 matches
Mail list logo
